CVS log for src/libexec/ld.elf_so/reloc.c

Up to [cvs.NetBSD.org] / src / libexec / ld.elf_so

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

Pull up following revision(s) (requested by riastradh in ticket #300):

	libexec/ld.elf_so/rtld.h: revision 1.145
	libexec/ld.elf_so/symbol.c: revision 1.74
	libexec/ld.elf_so/rtld.h: revision 1.147
	libexec/ld.elf_so/symbol.c: revision 1.75
	libexec/ld.elf_so/symbol.c: revision 1.76
	tests/libexec/ld.elf_so/t_hash.c: revision 1.1
	libexec/ld.elf_so/Makefile: revision 1.145
	libexec/ld.elf_so/Makefile: revision 1.146
	libexec/ld.elf_so/Makefile: revision 1.147
	libexec/ld.elf_so/reloc.c: revision 1.118
	distrib/sets/lists/tests/mi: revision 1.1280
	libexec/ld.elf_so/rtld.c: revision 1.215
	tests/libexec/ld.elf_so/Makefile: revision 1.21
	libexec/ld.elf_so/hash.c: revision 1.1
	libexec/ld.elf_so/hash.h: revision 1.1
	distrib/sets/lists/debug/mi: revision 1.409

The SysV ABI specifies that the symbol hash function should return only 32
bits of hash. Unfortunately due to an implementation bug and the fact that
the return type is unsigned long which is 64 bits in LP64, this can fail
in some cases: "\xff\x0f\x0f\x0f\x0f\x0f\x12". See:
    "https://maskray.me/blog/2023-04-12-elf-hash-function
From Ed Maste @ FreeBSD:
    https://cgit.freebsd.org/src/commit/?id=29e3a06510823edbb91667d21f530d3ec778116d

Need to write Unit Tests for this.

Oops wrong mask.

ld.elf_so: Split SRCS onto multiple lines.
Makes updates easier.
No functional change intended.

ld.elf_so: Sort SRCS.
No functional change intended.

ld.elf_so: Split hash functions into a separate file.
This way we can test them in isolation.
No functional change intended.

ld.elf_so: Add some known-answer tests for hash functions.

Make sure the testing mechanism detects the traditional overflow bug.

ld.elf_so: Split hash functions into a separate file.

This way we can test them in isolation.

No functional change intended.

Provide a stub _rtld_call_ifunc when a platform doesn't use one of the
common ones or doesn't provide its own.

Merge changes from current as of 20200406

Implement DT_GNU_HASH

DT_GNU_HASH serves the same purpose as DT_HASH, however it is a distinct
and faster apprach implemented and designed in the GNU toolchain in 2006.

DT_GNU_HASH is preferred whenever available.

Original GNU benchmarks claim 50% faster dynamic linking time.
https://www.sourceware.org/ml/binutils/2006-06/msg00418.html

Code based on FreeBSD and OpenBSD, both were based on DragonFlyBSD.

Implement and integrate GNU Hashing function

Define Elf_Hash struct that contains ELF/SYSV and GNU hash checksum.
Implement _rtld_gnu_hash() for DT_GNU_HASH.
Adapt existing code to compute all Elf_Hash types, instead of only
the ELF/SYSV one.
Rename _rtld_elf_hash() to _rtld_sysv_hash() to match the GNU toolchain
terminology.

_rtld_gnu_hash() uses Dan Bernstein's string hash function posted eons ago
on comp.lang.c.

Sync with HEAD

Synch with HEAD

binutils 2.31.1 can put copy relocations in the relro segment. Delay
protecting the relro segment for the main object until copy relocations
are done.

Sync with head

Move compatibility for pre-2.0 ELF binaries into separate file.
Early ELF binaries defined dlopen and friends in crt0.o by
using function pointers in the object handle passed from rtld.
This contract wastes space when many shared objects are allocated
and ties dynamic linker and rest of the system tightly together.
Fake the entry points in a static object and restrict them to the
platforms that had working ELF support at the time. Keep the magic and
version field used by modern crt0.o for all architectures. The checks
will be removed from crt0.o in a follow-up step.

Sync with HEAD.  77 conflicts resolved - all of them $NetBSD$

Rework ifunc support to address a number of short comings:
- Move to a shared _rtld_call_ifunc for rel and rela architectures
- Architectures using rel format must patch IRELATIVE non-PLT
  relocations like RELATIVE in additition to the later ifunc handling
- Consistently record the delta to the end of the relocation group for
  non-PLT IRELATIVE relocations

Hidden ifunc is now supported on all ifunc platforms, even when using
-fno-plt. The combination of -fno-plt and relro is broken due to
incorrect GNU ld output though.

Add IRELATIVE support for ARM, X86 and PowerPC.

Pull up following revision(s) (requested by uwe in ticket #1450):
	libexec/ld.elf_so/reloc.c: revision 1.110
PR port-macppc/47464 - Old binutils generated bogus zero-sized COPY
relocations for _SDA_BASE_ and _SDA2_BASE_ that happened to work by
accident.  They no longer happen to work when old binaries are run on
newer systems.  Explicitly ignore these bogus relocations.

Pull up following revision(s) (requested by uwe in ticket #1411):
	libexec/ld.elf_so/reloc.c: revision 1.110
PR port-macppc/47464 - Old binutils generated bogus zero-sized COPY
relocations for _SDA_BASE_ and _SDA2_BASE_ that happened to work by
accident.  They no longer happen to work when old binaries are run on
newer systems.  Explicitly ignore these bogus relocations.

Sync with HEAD - tag prg-localcount2-base1

PR port-macppc/47464 - Old binutils generated bogus zero-sized COPY
relocations for _SDA_BASE_ and _SDA2_BASE_ that happened to work by
accident.  They no longer happen to work when old binaries are run on
newer systems.  Explicitly ignore these bogus relocations.

Add support for GNU RELRO headers from Matthias Weckbecker.

- Print a warning for text relocations
- Don't remap the text segment executable while relocating

Catch up to -current, via patch, requested by christos in ticket #1126:

Add basic support for indirect functions. It allows providing a public
function symbol with an implementation choosen at run time.
Refactor calls to functions by address in ld.elf_so to create temporary
function descriptors on the stack, if the address is not leaked outside.

Limitations:
- no support for initialising static storage with function pointers
- no support for unnamed resolver functions

Inspired by FreeBSD's r228435 by kib@freebsd.org.

sync with head

Pull up following revision(s) via patch (requested by skrll in ticket #1724):
	rescue/list.ldd: revision 1.4
	lib/libc/dlfcn/dlfcn_elf.c: revision 1.7
	libexec/ld.elf_so/arch/mips/mips_reloc.c: revision 1.57
	distrib/sets/lists/comp/mi: revision 1.1512
	share/man/man3/Makefile: revision 1.56
	libexec/ld.elf_so/arch/mips/mips_reloc.c: revision 1.58
	usr.bin/ldd/ldd.c: revision 1.15
	libexec/ld.elf_so/rtld.h: revision 1.84
	share/man/man3/dl_iterate_phdr.3: revision 1.1
	libexec/ld.elf_so/rtld.c: revision 1.129
	libexec/ld.elf_so/arch/powerpc/ppc_reloc.c: revision 1.44
	libexec/ld.elf_so/rtld.h: revision 1.89
	libexec/ld.elf_so/arch/x86_64/mdreloc.c: revision 1.36
	libexec/ld.elf_so/map_object.c: revision 1.41
	libexec/ld.elf_so/arch/x86_64/mdreloc.c: revision 1.37
	libexec/ld.elf_so/arch/sparc64/mdreloc.c: revision 1.46
	include/link_elf.h: revision 1.10
	libexec/ld.elf_so/arch/i386/mdreloc.c: revision 1.29
	libexec/ld.elf_so/arch/vax/mdreloc.c: revision 1.26
	libexec/ld.elf_so/arch/alpha/alpha_reloc.c: revision 1.34
	libexec/ld.elf_so/arch/hppa/hppa_reloc.c: revision 1.31
	libexec/ld.elf_so/arch/alpha/alpha_reloc.c: revision 1.35
	libexec/ld.elf_so/Makefile: revision 1.94
	libexec/ld.elf_so/arch/hppa/hppa_reloc.c: revision 1.32
	libexec/ld.elf_so/Makefile: revision 1.95
	libexec/ld.elf_so/arch/arm/mdreloc.c: revision 1.31
	libexec/ld.elf_so/Makefile: revision 1.96
	libexec/ld.elf_so/arch/arm/mdreloc.c: revision 1.32
	libexec/ld.elf_so/reloc.c: revision 1.98
	libexec/ld.elf_so/arch/arm/mdreloc.c: revision 1.33
	sys/sys/exec_elf.h: revision 1.106
	libexec/ld.elf_so/rtld.c: revision 1.130
	libexec/ld.elf_so/load.c: revision 1.37
	libexec/ld.elf_so/rtld.c: revision 1.131
	libexec/ld.elf_so/load.c: revision 1.38
	libexec/ld.elf_so/rtld.h: revision 1.90
	libexec/ld.elf_so/headers.c: revision 1.36
	libexec/ld.elf_so/rtld.h: revision 1.95
	libexec/ld.elf_so/arch/i386/mdreloc.c: revision 1.30
	libexec/ld.elf_so/arch/m68k/mdreloc.c: revision 1.25
	libexec/ld.elf_so/symbol.c: revision 1.50
	libexec/ld.elf_so/symbol.c: revision 1.51
	libexec/ld.elf_so/arch/sparc/mdreloc.c: revision 1.43
	libexec/ld.elf_so/symbol.c: revision 1.52
	libexec/ld.elf_so/arch/sh3/mdreloc.c: revision 1.27
	libexec/ld.elf_so/symbol.c: revision 1.54
PR/39240: Satoshi Suetake: Don't fail when attempting to resolve weak symbols
when we are doing immediate binding, leave them alone and they will be dealt
with later during lazy binding. From skrll@
Implement negative cache checks for symbol lookups.
Uses the Donelist idea from FreeBSD.
Use alloca(3) instead of local xmalloc for creating our DoneLists.
This allows threaded programs to use us a little better, PR lib/43005.
Implement dl_iterate_phdr.
Somewhat taken from FreeBSD. Manual page from OpenBSD.

Implement lazy binding on hppa. rump_server needs it!?!?!

Mostly from OpenBSD.

Restore backwards compatibility by removing unnecessary addition of
dlvsym to Obj_Entry.

Add some comments.

PR/45015: ld.elf_so: support ELF symbol versioning
Applied latest patch.

Add support for DF_1_BIND_NOW, DF_1_NODELETE and DF_1_NOOPEN marked
objects, and the RTLD_NODELETE and RTLD_NOLOAD flags to dlopen(3).

Mark libpthread as DF_1_NOOPEN and use it to test the functionality.

Somewhat taken from FreeBSD.

Fixes PR 42029.

OK from christos and joerg.

sync to netbsd-5

Use fast_remainder32 for the ELF hash. For the hot cache case, this
speeds up Firefox startup by over 2% on AMD64.
Limit hash table buckets to 32bit.

Pull up following revision(s) (requested by skrll in ticket #1249):
	libexec/ld.elf_so/rtld.h: revisions 1.87, 1.88
	libexec/ld.elf_so/reloc.c: revision 1.101
Restore backwards compatibility for binaries referencing the main
Obj_Entry.

Remove cargo cult dlinfo entry point.

Put the dlopen,and friends entry points back.

They're needed by pre-2.0 binaries. Backward compatibiliity for these
was broken by another commit recently which I'll fix shortly.

Remove the entry points for dlopen() and friends. They haven't been used
since Aug 2003.

PR/39240: Satoshi Suetake: Don't fail when attempting to resolve weak symbols
when we are doing immediate binding, leave them alone and they will be dealt
with later during lazy binding. From skrll@

Pull up the following revisions(s) (requested by pooka in ticket #1088):
	include/dlfcn.h:	revision 1.20
	lib/libc/dlfcn/dlfcn_elf.c:	revision 1.6
	libexec/ld.elf_so/reloc.c:	revision 1.97
	libexec/ld.elf_so/rtld.c:	revision 1.125
	libexec/ld.elf_so/rtld.h:	revision 1.81
	libexec/ld.elf_so/symbol.c:	revision 1.48

Add Solarisa-like dlinfo() interface to the ELF dynamic linker.
Implement RTLD_DI_LINKMAP which returns a pointer to the linkmap
chain at the given object.  Other Solaris queries are currently
unimplemented.

Add Solarisa-like dlinfo() interface to the ELF dynamic linker.
Implement RTLD_DI_LINKMAP which returns a pointer to the linkmap
chain at the given object.  Other Solaris queries are currently
unimplemented.

Sync with wrstuden-revivesa-base-2.

Make sure to call dlerror() to clear any error messages after the relocations
are done and there are no errors.  This will prevent spurious undefined weak
symbol errors being returned by dlerror().

Fixes PR bin/39240.

Add LD_BIND_NOW support for mips.

This is the last arch to be done so remove the large ugly #if defined()
stuff in reloc.c.

Add support for immediate PLT binding on SH5. Compile tested only as
shared libs don't actually work on SH5 (yet).

OK'd by Steve Woodford.

Support immediate PLT binding on powerpc. Thanks to Simon Burge for
testing.

I can see the finish line...

Support immediate PLT binding on m68k. Tested on NetBSD/sun3 using the new
version of tme.

Not far now...

Support immediate PLT binding on vax. Tested on simh.

Nearly there...

Support immediate binding on x86_64.

Closes PR 30570. Thanks to TAMURA Kent for testing.

Support LD_BIND_NOW on sh3..

Support LD_BIND_NOW for alpha.

Add LD_BIND_NOW support for sparc and remove some relocation types that
are not supposed to happen in 32bit code.
From Nick Hudson.

Add LD_BIND_NOW support for sparc64.

Add __RCSIDs.

OK'd by mycroft.

Pull up revision 1.84 (requested by skrll in ticket #1726):
Don't print relocated content (in DEBUG mode) as it can cause
misaligned access. It needs to be done differently if needed at all.
Fix PR port-sparc64/20907.

Pull up revision 1.84 (requested by petrov in ticket #673):
  Don't print relocated content (in DEBUG mode), as it can
  cause misaligned accesses.  Fixes PR#20907.

Don't print relocated content (in DEBUG mode) as it can cause
misaligned access. It needs to be done differently if needed at all.
Fix PR port-sparc64/20907.

Apply patch (request by skrll in ticket #1702):
Bring "ld.elf_so" (mostly) in sync with NetBSD-current:
- MI and MD (e.g. under NetBSD-alpha) performance improvements
- RTLD_SELF, RTLD_NEXT, RTLD_DEFAULT support
- much better structured code
- closes PR bin/25464

Fix the dynamic linker for hppa using ELF.

Lazy binding doesn't work 100% of the time so force immediate binding.
One possible reason is that the PLT stub blows away r20 which the
compiler might not take into account.

Add a debug message.

Resolve dlsym(3) and friends directly so that dlsym(RTLD_NEXT,...) works.

Previously dlsym resolved to the version in crt0.o or libc which would
mean that the caller's shared object couldn't be determined correctly
using __builtin_return_address(0).

Mainly from FreeBSD, but adapted by me. Benefits of this solutions are:

	- backward comptibility maintained
	- existing broken binaries are fixed with a new ld.elf_so
	- __mainprog_obj can be removed from crt0.o
	- we do the same thing as FreeBSD

Fixes PR 22067.

OKed by Jason and Christos.

ANSIfy and de-__P().

Support immediate binding on arm.

Remove comment which is not true any more. alpha and i386 support
R_xxx_COPY relocations.

Back out the last two changes.  They were misguided, and the code worked
correctly before.

Now _rtld_relocate_objects() returns with error as soon as any of
subsequent calls in it fails.

Spotted by wiz and enami.

Simplify code a bit.

Remove the `self' args to _rtld_relocate_objects() and
_rtld_relocate_nonplt_objects().

Push the _rtld_bind() interface into MD code -- it's just a trivial wrapper
anyway.

Put back the `in_plt' nonsense for now.
It turns out there is some deep-seated wackiness WRT function pointer
comparisons...

Reverse the previous patch, but make the dbg()s more informative.

G/c dbg()'s.

Several small changes that shave 7-8% off the simple-exec-loop test:

* Rename _rtld_find_library() to _rtld_load_library().  It now calls
  _rtld_load_object() if necessary to actually load the object, rather
  than having the caller do it.  To do this, it also takes the `mode'
  argument that gets passed to _rtld_load_object().

* On a related note, remove _rtld_check_library(), and instead call
  _rtld_load_object() to instead try actually loading the object.  We
  save two extra namei's and a bunch of redundant work (almost
  literally the same code) this way.

* In _rtld_map_object(), mmap(2) the first page read-only, rather than
  read(2)ing it.

* In _rtld_symlook_obj(), compare the *second* character of the symbol
  name before calling strcmp().  (This first character is too
  frequently `_', and turns out to not be helpful, in libc.)

* Also in _rtld_symlook_obj(), remove the bogus STT_FUNC special case
  -- this also allows removing the `in_plt' argument to
  _rtld_symlook_list() and _rtld_symlook_obj().

Also:

* In _rtld_obj_from_addr(), rather than trying to look up `_end' in
  the each object, instead use obj->mapsize as the upper bound.

Make immediate binding work again by actually implementing
_rtld_relocate_plt_objects(). Note that this is for i386 only;
any efforts to make it work on other platforms are left to
those who are using them...

_rtld_bind() is not used on MIPS.  Should probably move it to MD code.

Nuke -DRTLD_RELOCATE_SELF and `dodebug' from orbit.

Pass down an additional flag to _rtld_relocate_nonplt_objects() which
indicates whether we're relocating ld.elf_so itself.  Use this in some places
rather than hackish tests on `dodebug'.  (The Alpha and HPPA `dodebug' tests
were actually noops, because RTLD_RELOCATE_SELF is not set, and therefore
dodebug is always true.)

Introduce a new flag, `isdynamic', which is used to remember whether the
executable was of type ET_DYN.  Use this instead of `mainprog' to determine
whether we need to do base-relative fixups of the PLT.  (This allows loading
non-relocatable objects, should we desire to do that at some point...)

More const poisoning.  Now the entire _rtld_bind() path useds a const Obj_Entry.

Split _rtld_relocate_plt_object() into two MD functions:
* _rtld_relocate_plt_lazy() fixes up all the relocs pointing to the PLT.  (On
  most platforms it just does a simple base-relative fixup; on SPARC it does
  nothing.)
* _rtld_relocate_plt_object() does immediate binding for a PLT entry.
The basic gist is that this saves a bit of time on SPARC (where the iteration
through the pltrela table was gratuitous), and a little less time on all other
platforms.  A whole lot of #ifdef'ed crap is moved out of reloc.c, too.

NOT tested on: hppa sh x86_64

A few things:
* Pass a symbol number to _rtld_find_symdef(), not a r_info.
* Don't try to do a symbol lookup when we find an unsupported relocation;
  instead get the symbol name from the referencing object's strtab.
* Add preliminary support for `-z combreloc'-style startup optimization on
  i386, `#ifdef COMBRELOC'.

Rename _rtld_relocate_nonplt_object() to _rtld_relocate_nonplt_objects(),
and push the outer loop into it.  This actually shaves a couple % off startup
time at least on PCs.

Split _rtld_relocate_nonplt_object() into separate MD files.

Split MIPS relocation goo into mips_reloc.c.

Simplify a couple of tests to !__mips__.

Combine the code that skips COPY relocs on most platforms (other than MIPS).
Remove the code that skips JMP_SLOT relocs (thus making ld.elf_so bomb if it
hits one in the non-PLT relocation table).

Remove unused arguments of _rtld_find_symdef().

Remove support for PowerPC static relocations, which are not present in shared
objects, and in any case are implemented WRONG.  (They're symbol-relative, not
base-relative.)

Make _rtld_setup_pltgot() a consistent interface on all platforms.

Add support for dynamic linking on HP PA-RISC.

Add support for dynamic linking on the SuperH.  From Marcus Comstedt,
with some bug fixes (namely, don't forget to consider the addend when
processing relocations) from me.

Do not attempt to relocate any pltrela entries that are not JMP_SLOT on sparc64.

Small whitespace fix.

Upon further investigation, do NOT include an addend in RELATIVE
relations (the addend should always be zero, but this is safer).

Include the addend in GLOB_DAT relocations (!!).

* Move The Alpha R(RELATIVE) handling into an Alpha-specific code block,
  thus isolating the "iffy hueristic" from the rest of the relocation code.
* In the "iffy hueristic", use _GOT_END_, not _DYNAMIC.
* Include the addend in Alpha R(RELATIVE) relocations.

Add code to peek into the PLT and determine which PLT entry format
is being used.

Supply two lazy binding routines for Alpha: one that works with the
old PLT format, and one that works with the new.

XXX We currently always use _rtld_bind_start_old() in
_rtld_setup_alpha_pltgot().  We need to add code to peek
into one of the PLT entries to see which format it's in
and pick the correct binding routine.

Check in workaround for BFD ABI change on MIPS: the new MIPS BFD generates
relocs for local symbols as section-relative REL32 relocs with values based
at 0, whereas previously BFD generated those relocs with values based at the
start address of the section.

This code adjusts all STT_SECTION REL32 relocs which have values less than
the start of the section by adding in the base address of the section.  This
may limit section sizes to 2GB, but that shouldn't be a problem for now.

XXX: Needs to be followed-up with binutils list to get closure on which
interpretation is correct (the new ABI or the old one), I just haven't
yet had the time to chase that.

Pull up revisions 1.34,1.41-1.42 (via patch, requested by skrll):
  Add init/fini section support in crtbegin and crtend, and introduce
  support for DWARF2 exception handling.  Fixes PR#12865, PR#13488,
  PR#13489, and PR#13491.  Also fix ld.elf_so to deal appropriately.

Reorganize PowerPC cold a little.  The return value from _rtld_bind_pltgot()
was ambiguous in the case of a weak symbol that was not defined.  This caused
RTLD_NOW to fail badly with shared libraries linked against the new crtbegin.o.

Workaround for binutils 2.11 behavior.

Fix ABS32 ARM reloc case.  ELF shared libs on ARM now work!

Bugfix for _rtld_start.  Enable common plt processing code for arm

Enable arm if OBJECT_FMT is ELF.  Initialize first few GOT entries for ARM.

Add support for ARM ELF dynamic relocs.

Preliminary x86-64 support. Not fully functional yet (toolchain bugs).

Pull up revision 1.35 (requested by kleink):
  Rename ElfNN_RelA to ElfNN_Rela, which is what SVR4 ABIs and the
  rest of the world uses.  Fixes PR#12274.

Sync with <sys/exec_elf.h> rev. 1.45: ElfNN_RelA -> ElfNN_Rela.

avoid nested extern

Pullup 1.33 [dan]:
format warning fixes from he, after testing on 64-bit elf (alpha)

format warning fixes from he, after testing on 64-bit elf (alpha)

Apply patch (requested by he):
  Print possible 64bit values with %ld and cast to u_long.

Simplify some stuff.

Approved by thorpej:
Add support for sparc64, including MD code and variable page size support.
Modify some variable types to insure they are the correct width on sparc64.

basesrc/libexec/ld.elf_so/arch/sparc64/Makefile.inc	0 -> 1.1
basesrc/libexec/ld.elf_so/arch/sparc64/mdreloc.c	0 -> 1.3
basesrc/libexec/ld.elf_so/arch/sparc64/rtld_start.S	0 -> 1.2
basesrc/libexec/ld.elf_so/headers.c			1.6 -> 1.8
basesrc/libexec/ld.elf_so/reloc.c			1.26 -> 1.27
basesrc/libexec/ld.elf_so/reloc.c			1.29 -> 1.31
basesrc/libexec/ld.elf_so/rtld.c			1.34 -> 1.35
basesrc/libexec/ld.elf_so/rtld.c			1.36 -> 1.40
basesrc/libexec/ld.elf_so/rtld.h			1.26 -> 1.27
basesrc/libexec/ld.elf_so/symbol.c			1.7 -> 1.9

Catch up with type changes.  Also:
* Apply DT_PLTRELSZ to (one of) pltrel or pltrela *after* we've finished
  parsing the headers, so we know which one.
* Fix sparc64 bogons.  (It works now!)

Fix some issues with SPARC V9 ELF binaries.

More VAX ELF changes.

Update to reflect VAX ELF reality.

Added support for the SPARC v9 ABI.

PowerPC: support ADDR16_LO, ADDR16_HI and ADDR16_HA relocations.

add hooks to support the dladdr function.

Pull up to last week's -current.

Use "addend" for R_68K_GOT32, R_68K_PC32, and R_68K_32 relocations.
I'm not sure whether R_68K_GOT32 relocations appear in shared
objects, though....

Example:
	lea	%pc@(got+0x1234@GOTPC),%a0	| R_68K_GOT32
	bsrl	foo+0x5678			| R_68K_PC32
	lea	bar+0x9abc,%a1			| R_68K_32

Changes from msaitoh to fix local/global symbol confusion, and to fix weak
symbol support -- updated by me for elf.h changes.

More SVR4-style ELF cleanups.

Initial VAX ELF32 support (mostly untested).  It's checked primarily for
testing and archival for now.  I don't expect anyone to work with it
since the binutils and gas changes are still pending.  But you got to
crawl before you walk.

Add m68k code.

Fix i386 problem, when we are relocating ourselves... We cannot depend
on the address of GOT or _DYNAMIC.

Make this compile without -DRTLD_LOADER and fix printf format problems on
the alpha.

Not all ports use 'tmp'.

KNF

Fix typo introduced in rev 1.11

Introduce pre-preocessor symbol RTLD_INHIBIT_COPY_RELOCS.
Currently, only mips does not include COPY_RELOCs code.

Completely construct the first PLT entry and correctly compute the
PC relative offset in the call insn.

Partial fix for Alpha breakage.  s/tmp_value/tmp/ in two places, and don't
try to check _rtld_objself when doing R_TYPE(RELATIVE) relocs, as the
Alpha ld.elf_so contains them!

I'm not sure the logic for the RELATIVE fix is quite right; it happens to
Work on alpha, but do we actually need to make sure we aren't relocating
ld.elf_so there?

XXX: I am working on making use of RTLD_RELOCATE_SELF on alpha, which may
make this hack of a fix moot.

Cleanup global variable access so rtld can be self relocating. Also KNF
reloc.c and move md sparc parts to arch/sparc/mdreloc.c
Now we don't need to load at address 0 on the i386 and the mips should be
trivial to fix.

Fix glitch in previous.

Add an approximation of the necessary sparc bits in here.
I've not followed the trend in this file by trying to wrench the MD
code into individual functions. Rather I,ve replaced several functions
wholesale.  Anyway, this whole file needs to be re-done.

tv is right; we don't need to specialcase the i386.

On everything but the i386, don't write new values into the GOT that are
not pointing to the PLT in _rtld_relocate_plt_object().  I'm not quite
sure why the i386 added the extra relocations, but it removed the `return
0' needed for other arches.

THIS DOES NOT YET FIX BUILDING LD.ELF_SO ON ALPHA.  There are some other
binutils 2.9.1 related issues, but this does return build ability to
2.8.1 based as and ld.  Will do more investigation later.

Fixed for i386/elf.

Fixup some printf format argFixup some printf format argFixup some printf format argFixup some printf format argFixup some printf format argFixup some printf format argFixup some printf format argFixup some printf format argFixup some printf format args.

Add i386 bits

Add powerpc support.

Sync with trunk, per request of mycroft.

Add support for mips.

First cut at an ELF shared loader.  Originally from John Polstra's FreeBSD elf
kit, then hacked on by Matt Thomas <matt@3am-software.com>, then by me (to
make it work with new versions of the toolchain, etc.).  This runs, but it's
in serious need of cleaning and/or a fair bit of reworking.  See the README
file for more information, and a list of things to do.